Mood disorders and suicide in youth represent major health concerns. Clinicians working with these youth need effective intervention strategies. However, there are no well established psychosocial interventions for prepubertal children with depression or manic-depression. Existing literature suggests that reducing expressed emotion (EE) via family psychoeducation improves outcome for adults with mood disorders, but similar studies in children are lacking. We developed an eight session, manual-driven, multi-family psychoeducation group therapy program (MFPG) designed to serve as an adjunct to the ongoing medication management and individual/family psychotherapy a child receives. Preliminary studies suggest MFPG is associated with multiple improvements for families, including: increased knowledge about mood disorders and improved coping skills in parents; improved family climate; increased perceptions of social support for parents and children; and increased effective treatment utilization. However, these studies have had significant methodologic limitations. In the proposed study, we hope to rigorously test the efficacy of MFPG with 165 mood disordered children aged 8 to 11. Participants will be recruited from multiple settings to obtain variability in socioeconomic status (SES), ethnicity, prior treatment history, and typical access to mental health services. Children and their primary and secondary parents/caregivers will complete pre-treatment assessment batteries at Time 1, then will be randomized into immediate MFPG plus treatment-as-usual (TAU) or a wait-list condition (WLC) plus TAU. All will be reassessed at Time 2 (approximately 3 months after study entry), Time 3 (7 months), and Time 4 (12 months). After the Time 4 assessment, WLC+TAU families will receive MFPG. A final assessment (Time 5, 15 months) will occur post-treatment. Hypotheses are: 1) MFPG+TAU families will function better than WLC+TAU families at Times 2, 3 and 4; 2) all families will improve functioning from immediately pre-treatment to immediately post-treatment; 3) several baseline variables (healthier child and parent functioning, higher caregiver concordance regarding treatment, and greater access to services) will be associated with better outcome for all participants; and 4) differences in gender and ethnicity will be unrelated to outcome.